Sovtek Mig-50h : Filter caps fried, fuse blown... what else?

[williestargell]

I'm working on a Sovtek Mig-50H Tube Midget w/ master volume.

The guy had the input power voltage selector set to 220v, using a 120v outlet. It stopped working for him, blew a fuse, but still powered on. There is a loud hum if it is on.

I looked inside and the big 220uF 300v filter caps are obviously burned and freed themselves from solder connection to PCB. There's a 470ohm 6.5v wire wound resistor that is fried too.

So I guess I'm asking if the switch being set wrong can cause this, and what else to look out for. I will retube and bias, and touch up a few frayed wires. The caps look original, khaki colored "IC" branded. A yellow wire from the power transformer is right in there with all the burnt nonsense.

They use a 4amp and 2amp ceramic fuse, though I haven't found any in the smaller size. would a slow-blow work, or a regular fast-acting fuse? I know all the electrolytic caps need replaced eventually...

Any known causes of filter caps frying?

Any help is greatly appreciated!

ps: The voltage selector switch is a bit of a mystery. It has 5 positions it clicks to, but only 4 labeled. from top to bottom there is 100v, 120v, 220v, 240v, and the mystery unmarked position at the bottom.

 

[andyfromdenver]

i don't know what it looks like inside those amps, but i think it would be worthwhile to disconnect the PT and test it, before putting it in circuit again. The multivoltages on a switch sounds like a major pita too. who knows what they did to it.
is there a schematic for the power section?

 

[Wally]

That sounds like a big job there. PITA, deeeeeluxe. +1 on testing the PT for continuity first and voltage output secondly. ...and then you still don't know what it will do under load, right? IF the PT looks 'good', then take a look at the OT's continuity. IF it looks good, recap the amp.....filters and biasing caps. Have fun with that....check all of those PCb traces near that 'fried zone' for continuity.
IF you get that far, fire the amp up with known good tubes. IF you have a current limiter, use it. It will indicate shorts. IF you don't have one, google 'current limiter circuit' and build one...simple to do.....light bulb and socket, AC receptacle...box to put it in. A block of wood worked for mine. (;^)

As for "Any known causes of filter caps frying?"
In this case, one might deduce that the cause was 'operator error', right?

 

[CoolBlueGlow]

What WALLY says...as usual.

A few expansions too...

Step one - get the correct schematic. Here is a start. Probably not exact, but better than nothing.

http://elektrotanya.com/PREVIEWS/63463243/23432455/egyeb/sovtek_mig-50.pdf_1.png

Step two - get the real story. Sorry, but 120 on a 220 primary should not cause the failure(s) you describe. I have deliberately connected 220 wired primaries to a 120 supply to observe the failure mode. The amp will not operate, but no component failures typically result.

Re-interview the owner. Ask him "Is he sure he didn't set it to 100 volt?" Why? Simple...THAT error would definitely cause the primary filter caps to fail, since they are typically operated so close to failure voltages anyway.

This amp has the five way mains transformer. 100 volt is for Japan.

Power a 100v mains setting at 120v means a 20v error on the primary - which is multiplied by the ratio of the secondary winding, which means the caps would see 70-100 volts more than their rating, which translates into a failure suspiciously similar to what you've described.

Suggested repair approach...others chime in. Yes, I am ASSUMING you are a competent tech in the general sense of the word. If not STOP - TAKE IT TO A TECH.

Study the schematic to familiarize and discover any differences in the amp you have on bench. Once comfortable, proceed. TUBES OUT.

Check all tubes for function.

Next, replace ALL SEVEN primary power supply caps. 2@220uf/350 the next four are pairs at 47UF/350v. Uprate them to 450v. Uprate that final solo 47/450 to 47/525.
Replace bias circuit filter cap.
Replace the burned 470 ohm resistor you describe, which MAY BE associated with the bias supply circuit (or MAY BE the 500 ohm 3w resistor in node one of the PS.) figure it out.
Replace the MOV on the primary, just past the 3 amp primary fuse.
Check/replace rectifier diode bridge. It might be torched.
Check/Replace the 500 ohm 3 watt resistor at node one of the power supply. (If that 470 wire wound you mention actually was the bias supply 470 shown on the schematic)
Check/replace the 5.1k and 51k 1w at nodes two and three of PS
Check/replace the 3kv plate to ground diodes.

Once you've done this, obviously still TUBES OUT, Set mains transformer to 120v (!) Disconnect the four power supply node feeds at their appropriate points. Power up the amp's power supply ONLY using a variac and measuring current. If current and voltages it look fairly normal...as in near zip with no tubes and no load, then check filament voltages. Tubes out, they will be a little high, but should be close. 6.7-7.2vac would be typical.

Reconnect power supply node one (ct of the OPT) and variac it up again - still no tubes in. Look for normal current levels for what should still be a no-load condition.

While powered up, check bias supply function - set to schematic recommended voltage. Check ripple. If good, shut her down, install power tubes ONLY, and connect node two, (G2 supply) and re-variac as per usual, carefully watching current. Should now show a nominal but definite load. Check filament v again. Should be close, but perhaps a very little above 6.3vac. Check plate voktages...should be close, but perhaps a little high. Check for excessive ripple.

If output tubes' PV and G2v is stable and normal, connect nodes three and four using similar approach. You should be getting close to ready to rock.

From here on it will be just chasing gremlins, presuming the above makes sense.

IF IT DOESNT, TAKE IT TO A TECH.

(Did I mention that if you are not a tech that you should take it to a tech?

CBG

 

[Wally]

CBG, I couldn't see how that voltage 'mismatch there would cause these problems either....but electrons are strange and independent beings...beyond my comprehension at times. A mismatch into that 100 primary winding is more likely, I agree.

From what I see on that schematic, those fitler caps are arranged in 3 pairs in series...so all of those voltage ratings are good to go, aren't they, CBG. The first pair...2 x 220/35V yields 110mfd of capacitance at 700V. The second and third stages are each made up of a pair of 47mfd/450V caps in series for a capacitance/voltage rating of 23.5mfd/700V...again, good to go....or they were until they were subjected to whatever they were subjected.

Great layout for an approach to the repair, CBG....and great advice.

 

[CoolBlueGlow]

Yes, agreed, and I saw that when checking out the schematic. so not really clear why over voltage would wreak such havoc. Still, the OP description makes the first two caps sound fried, so something had to make them over-volt, right?

That makes me suspect some failure UP the line, which would hint at similar failures down the line, since the circuit design for nodes two and three is so similar. (I mean paired caps)

Possible that the rectifier shorted due to the impulse overload and then took a cap with it. (Amp has a standby switch) If ONE of the paired caps shorted, that instantly exposes the second cap to the full voltage across the load.

Either way - old caps, and at least two have failed...and it has something to do with someone monkeying around with wrong PT voltages. Seemed a prudent recommendation to replace all.

But you are right - a curious and indirect path to explain the failure...just hints of troubles.

 

[CoolBlueGlow]

Oh, here it is Wally!

A quick Google search finds numerous folks discussing MIG50 plate voltages commonly as high as 640 VDC!

There you go. 120VAC connected on the 100V primary results in 768vdc on the rectifier's output. Instant failure, even with 700 volt cap pairs.

Amp has a standby switch at the rectifier...BEFORE the filter pack. Power on, no problem...but flip the standby switch and the power supply section fails in seconds.


That explains it.

 

[williestargell]

You guys are amazing with such detailed info. Wally, CoolBlue, thanks for the suggestions and walkthrough!

I had been imagining standard ancient cap failure when he was describing it to me. The chassis hasn't been screwed down into the cab properly for years and gets banged around a lot, who knows what that switch was set to or if it was even off of 120v. (i'm making him a plate for the back of the head as a favor)

We don't have a variac here, I'm going to do all the physical parts replacement I can, then have a tech test the PT before plugging in. It's one of those things where the diagnostic and labor rate is probably more than the amp is worth, but he can drop $60 to test the PT after the clearly faulty or obviously needed parts are replaced.

I got a little more info from him, he says it powered down completely while in standby. He replaced the 4A ceramic fuse with a regular 5A fuse, it powered on and power tubes were glowing. There was a loud hum, no input signal. Preamp tubes were covered, so he didn't see if they were alight.

 

[Wally]

640, eh? That is above the norm. OF course, I have been working on a 410HD130 Music man that has 730 on the plates with 63% of MPD. IT had 745 there when I started. (8^O...of course, MM put the screen voltage much lower....395VDC or so. that saves the screens, and as I understand it, 70vdc is not to far out of the max for EL-34's...700 as spec'd.
640 volts for a 6L6 is pushing things. I wonder what the screens are seeing?

RE: that multitap primary PT....unless the owner is traveling the world, I might suggest that hardwiring the 120 tap might be a good idea. That prevents any future misapplication of power.
Willie, CBG has given you theh plan....good learning procedure.

 

[williestargell]

they are touring europe this summer, but i'm going to convince him to rent a head. this thing would disintegrate. hardwiring 120v sounds like a great idea!